Chapter

Documenting the Ohio Hopewell

Mortuary Record:
The Bioarchaeological Data Base
D. Troy Case and Christopher Carr

This chapter introduces the reader to the data

base of Ohio Hopewell burials and associated
grave goods that have served as the foundation for
reconstructing the social and ceremonial organization of Scioto Hopewell peoples and other
Ohio Hopewell groups, as presented in Part II of
this book and in Gathering Hopewell: Society,
Ritual, and Ritual Interaction (Carr and Case,
2005c). The data base, called HOPEBIOARCH,
describes the tomb characteristics, artifact inclusions, artifact positions relative to the body, and
age and sex estimates for 936 burials representing 1483+ individuals from 112 mounds and
burial areas in 50 excavated Ohio Hopewell sites.
It also describes the contents of 77 ceremonial
deposits from 47 mounds or areas in 19 sites.
The total numbers of sites and mounds with
burials and/or ceremonial deposits are 52 and 126,
respectively. A ceremonial deposit is defined here
as a collection of several artifacts, of the same
type or different types, that appears to have been
intentionally placed together and buried without
accompanying human remains. A ceremonial
deposit can also be a single artifact found in a
specially prepared area (e.g., a crematory basin,
the horseshoe-shaped feature at the North Benton

site). Accumulations or deposits containing only

faunal or other organic elements, fragments of
a single artifact type that appear to be utilitarian (e.g. plain pottery), or both were not
included in the database (see Chapter 8). The data
base includes burials and ceremonial deposits
from all excavated Hopewell burial mound
sites in Ohio that have been inventoried by
Seeman (1979a:262, Table 2) and Fischer (1974:
359362, Appendix A4), that can be shown to
fall within the Hopewell tradition by modern
criteria, and for which intrasite provenience information on burials is available. Several additional
small Hopewell sites are also included in the
database to round out the inventory of graves
from identifiably Hopewell sites in Ohio. These
smaller sites are: Dayrs Farm mound, Finney
mound, Fortney mound, Glen Helen mound, Lee
mound, Manring mounds, Martin mound, Pence
mound, Perry Township mound, Shumards
Farm mound, Snake Den mound group, Stone
mound, and Yant mound. We believe these
represent all published and unpublished Ohio
Hopewell cemeteries that have been excavated
and for which written documentation exists in
museums, historical societies, and universities.

335

336

The purpose of this chapter is to describe

the structure of the HOPEBIOARCH data base
generally, the rationale behind this structure,
the mechanics of its creation, the kinds of
information sources used, and the locations
of both descriptive and complementary information within this book.

THE DATE BASE

AND ITS DOCUMENTATION
The HOPEBIOARCH data base, itself, is on
the accompanying CD. It is presented in four
different formats (Appendices 6.1AD). Two
are EXCEL files. One contains the data base in
its full form, with provenience, demographic,
grave good, tomb form, and grave good
positioning information. The second EXCEL
file omits the bulky information on grave good
positioning and thereby makes quick scanning
of the matrix for all other information much
easier. The remaining two versions of the
HOPEBIOARCH data base have the same
contents as the first two, but are tab-delimited
files.
Most aspects of the data base are described
in Chapters 710. These chapters define its
variables, cases, and contents, and provide
information that will assist researchers in
designing future studies that make use of these
data. The archaeological sites covered in the
HOPEBIOARCH data base are described in
Chapter 7. Information is provided about site
location, including the nearest township, the
major and minor drainages in which the site is
found, and the locations physiographic characteristics. Distances from various towns are for
those towns that existed and were reported at
the time of excavation. Details of site size and
form are also provided, including the presence
or absence of earthworks and the numbers of
mounds or burial areas identified. These data
can be used to help organize the various sites
into groups by geographic and cultural region,
and by size and function to a degree. Other
information provided in Chapter 7 helps to
define the quantity and quality of data available
for each site. Included are estimates of the

D. TROY CASE AND CHRISTOPHER CARR

extent of excavation of burial areas, and the

quality of reporting of details such as the
ages and sexes of skeletons, the stratigraphic
and horizontal locations of burials at the site,
and artifact locations relative to each skeleton.
In addition, a bibliography is presented that
lists published and unpublished archaeological
sources of information on each site and on the
ages and sexes of individuals buried at each site.
Chapter 7 also introduces Appendix 7.2, which
contains 101 maps that show the internal spatial
layouts of 64 mound floors and 14 ceremonial
enclosures or mound groups. This appendix is
included on the CD. The maps, combined with
information from the data base, can be used
to explore the spatial distributions of mortuary
traits across charnel house and mound floors.
An overview of the 545 variables found in
the data base and their corresponding variable
states is given in Chapter 8. These variables
are divided into three types: 177 primary
variables that describe particular artifacts, grave
attributes, provenience identifiers, and etc.,
74 numeric variables associated with some
of these primary variables, and 294 position
variables that indicate the location of artifacts
within graves and relative to the skeleton or
cremation. Definitions of each primary variable
and burial state code are provided, along with
general descriptions of the numeric and position
variables. Figures in Chapters 14 that depict
one or more examples of artifact types are cited
in Chapter 8 to help clarify the artifact definitions. The artifact classification system used
in the data base distinguishes items primarily
by their formal and material qualities. It also
attempts to capture the social and ceremonial
functions of the artifacts. Thus, for example,
copper and mica cutouts are treated as separate
variables, while the various forms of these
cutouts are richly described in a variety of
different variable states. Most of the variables
in the data base are mutually exclusive of
one another. However, in a few cases, certain
variables that overlap with each other were
created for specific analytical purposes. Such
cases of redundancy are clearly identified in the
variable descriptions so that other researchers
may recode the data, if necessary, to suit their
own analytic purposes.

DOCUMENTING THE OHIO HOPEWELL MORTUARY RECORD

One suite of variables in the

HOPEBIOARCH data base that is absolutely
essential for sociological reconstructions is the
estimated ages and sexes of individuals and the
reliability of these estimates. Age and sex data
were gathered on Ohio Hopewell skeletons
over the course of 120 years by different
researchers with greater or lesser experience
and using a variety of different methods. The
reliability of the age and sex information
available for human remains from several of
the Hopewell sites in the data base is addressed
in Chapters 9 and 10. Chapter 9 assesses the
data available as of 1998 for the Ater, Esch,
Harness, Hopewell, Rockhold, Seip, and Turner
sites, and draws conclusions about which
assessments can be used in social analyses
with relative confidence, and which should
be treated with caution. Several appendices to
Chapter 9 are located on the CD and list the
specific age and sex assessments available from
all known sources. These data are also coded
within the data base itself as several different
variables. Chapter 10 describes very recent
age and sex assessments that have been made
for skeletons from specifically the Hopewell
site and that use a wide array of osteological
and dental techniques and two multivariate
approaches. These new approaches have added
to the number of skeletons from the Hopewell
site with reliable age and sex information, and
have refined many of the previous assessments.
Two important appendices from this chapter,
Appendices 10.3 and 10.4, are found on the
CD and provide a provenience by provenience
account of the information available on each
of the skeletons encountered by Warren
Moorehead and Henry Shetrone during their
excavations of the Hopewell site. The appendices weave together information from site
reports, field notes, and the skeletal collections
themselves. The appendices describe which
skeletons were collected, whether the bones
exhibit cutmarks and copper staining, and
include detailed descriptions of culturally
modified human remains from the site. The
specific bones curated for a particular skeleton
are also sometimes recorded, especially for
burials from which only a few bones were
collected.

337

CONSTRUCTING
THE DATA BASE
The data base was assembled by the authors
in a number of overlapping stages over a
period totaling approximately eight years. We
were assisted in this task by a number of
graduate students without whose help the data
base might not have been completed. We
began our work of documenting Ohio Hopewell
burials with the larger sites of Hopewell, Seip,
and Turner. These sites were targeted because
their reports were published and available,
and contain detailed descriptions of individual
burials and ceremonial deposits. Our approach
at this initial stage was to read through the
site reports, gather together all relevant information about each burial described in various
portions of a report, and then to write a
bulleted summary of the nature and contents
of each grave and ceremonial deposit using
the original terms that the excavators and
authors had used for the artifacts and tomb
forms. We did not boil down their descriptions
into a priori descriptive classes. These detailed
summaries of graves and ceremonial deposits
came to be called provenience sheets. They
are reproduced in Appendix 6.2. In addition,
the three sites were selected because they were
known to encompass much of the spectrum
of artifact classes found in Ohio Hopewell
mortuary sites generally (Seeman 1979a). Our
first pass through these site reports helped us
to define the types of variables that should be
present in the data base, the kinds of information
consistently reported by excavators versus that
which was idiosyncratic, and what additional
information and forms of documentation would
be necessary if the data base were to be useful
for conducting intrasite and intersite mortuary
analyses.
Once we had secured an understanding
of the diverse kinds of information commonly
recorded for an Ohio Hopewell mortuary
site, we expanded our coverage to include
burials and ceremonial deposits from other
Ohio Hopewell sites that were published.
We also noted additional sites that Seeman
(1979a) and Fischer (1974) listed but that
had only unpublished reports available, and

338

only in museums or historical societies. The

second stage of our data collection efforts
involved grant-funded research trips to examine
unpublished field notes, site maps, accession
records, field photographs, and some of the
artifacts from published and unpublished Ohio
Hopewell sites. Sources of these data include
the Ohio Historical Society in Columbus, the
Field Museum of Natural History in Chicago,
The Peabody Museum of Archaeology and
Ethnology at Harvard University, Hopewell
Culture National Historical Park in Chillicothe,
Ohio, as well as a number of smaller public and
private collections (Chapter 7). The examined
records contributed many new proveniences
for sites that are documented only in part in
published site reports, filled out information on
burials and ceremonial deposits described only
to a degree in the reports, and added many new,
unpublished sites to the project.
The examined records highlighted many
inconsistencies among the data sources. For
example, field notes written by the excavators
sometimes disagreed with their own site reports
about the specifics of certain burials. Inconsistencies such as these were assessed and a
decision was made on a case by case basis
as to which information source to trust. We
tended to give the greatest weight to the field
notes and maps when discrepancies involved
numbers of individuals in a grave and positions
of artifacts in a grave. We were more likely to
follow the site reports for descriptions of the
specific nature of artifacts and their numbers,
under the assumption that the artifacts might
have received closer scrutiny at museums or
by other experts prior to publication. Museum
accession records were also used to assist with
clarifying the types and numbers of artifacts
present in particular graves. However, information from the accession records about artifact
numbers was treated conservatively because it
was not always clear, for example, whether
broken specimens were counted by the number
of total objects represented or the number
of pieces present. When inconsistencies arose
between accession records and site reports
or field notes, decisions about which information to use were again made on a case

D. TROY CASE AND CHRISTOPHER CARR

by case basis, sometimes taking into account

what we knew about the tendencies of the site
excavators and reporters. Our guiding principle
was to maximize the specificity of the data
available while minimizing the probability that a
particular artifact or burial characteristic would
be included in the description of a burial by
mistake. Our bias was to not include information if our best assessment was simply
that the artifact or characteristic was possibly
present or probably present.
The third stage of our building the
HOPEBIOARCH data base involved defining
its variables from the descriptions that
excavators and site reporters had provided and
that we had summarized in the provenience
sheets. These definitions were then used to
develop the data base structure and to code
information into the data base. This stage was
begun after the provenience sheets for the
first few sites were created from published
site reports, but before additional, unpublished
information from museums had been reconciled
with them. The provenience sheets and data
base were updated several times as new sources
of archival information were tapped.
Our aim was to create a data base that
would contain all of the data available for
each burial at a site, while being structured in a way that would facilitate social
analyses within and between different Ohio
Hopewell sites. Therefore, the 177 primary
variables in HOPEBIOARCH are ordered by
analytical categories useful in sociological
mortuary studies. The biological categories
concerned with ages, sexes, and numbers of
individuals in a grave are found near the
front of the data base, followed by variables
that consider burial characteristics such as
tomb form and grave dimensions. Artifact
classes appear next, organized into broad suites
of classes, primarily by the social roles in
which they were inferred to have been used
(Chapter 11) and secondarily by form and
raw material: the paraphernalia of shamanlike practitioners, other ceremonial equipment,
the paraphernalia and role markers of nonshamanic leaders and other important people,
clan markers, items of wealth and personal

DOCUMENTING THE OHIO HOPEWELL MORTUARY RECORD

decoration, utilitarian objects, and fancy raw

materials. Each artifact class is accompanied
by three variables that together describe for
a burial the position of artifacts of that class
relative to the corpse or cremation, where
information on artifact position is available.
Most of the variables are descriptive of artifact
classes or burial characteristics, but a few are
interpretive. For example, the variable water
barriers describes any set of artifacts or natural
materials that were placed around a grave apparently to act as a water barrier to ghosts, much
like the water barriers that were constructed
around some Adena mounds (Hall 1976b; see
also Carr 1998, 1999a, 2000b). Materials that
might signify a water barrier are those that
come from water or have a color or shine that
might represent water, such as mica, shells,
pearls, limestone, and light colored rocks (e.g.
Figures 5.3AE). Quite a few examples of
graves surrounded by such materials are known
from Ohio Hopewell sites, making the water
barrier a significant interpretive variable for
mortuary studies.
The original version of HOPEBIOARCH,
upon which many of the analyses in Gathering
Hopewell (Carr and Case, 2005c) were based,
was completed in 2001. This version of the data
base included information on 854 individuals
from 33 Ohio Hopewell sites, as well as 65
ceremonial deposits from 14 sites. The data
base has since been expanded to include a
total of 1052 individuals and an estimated 431
commingled human remains from 50 sites, as
well as 77 deposits from 19 sites. These sites
encompass, as far as we know, all Hopewell
mortuary sites in Ohio for which written information on internal provenience is available.
A few new variables have also been added
to this most recent version of the data base.
Most critical, revisions have been made to
the age and sex assessments of some burials
from the Hopewell site (see Chapters 9, 10).
These estimates were not available at the time
that studies were being made for the book
Gathering Hopewell. The particular modifications made to the age and sex data can be found
in Table 9.2.

339

PROVENIENCE SHEETS
Appendix 6.2 contains the provenience sheets
for all burials coded in the data base. Each
sheet is a bulleted list of the characteristics
of a burial or ceremonial deposit of artifacts,
and was compiled from one or more sources
of information. Each sheet served as a transitional step in coding the burials and deposits
into the data base. A sheet specifies the
type of provenience (burial or deposit), the
primary source of information on the provenience, and its form and size (e.g., tomb form).
For burials, this information is followed by a
brief summary of the human remains, including
an indication of burial type (inhumation vs.
cremation), number of individuals represented,
and other relevant details about body position,
estimated stature, head orientation, etc., when
these were recorded in documents. For both
burials and deposits, a list of the types of
artifacts recovered and their numbers follows.
Each artifact type is described in as much
detail as was necessary to create an appropriate
code for its inclusion in the data base. Typical
descriptions include the material from which the
item was made, some indication of its absolute
or relative size, and its location in the grave
relative to grave features or the human remains.
In cases where the excavator specifically
mentioned that no artifacts were recovered,
this is also noted. For many proveniences,
excavators did not explicitly say whether or
not artifacts were recovered, so the provenience
sheet may simply list a skeleton without any
indication of whether artifacts were associated
with it. For a small number of proveniences,
photographs of them during their excavations provided some additional information
on tombs, artifacts, and spatial layouts. Such
information, when present, is indicated under
the heading Photo, below the description
of artifacts.
The provenience sheets found in
Appendix 6.2 complement the data base
in several useful ways. First, they allow
researchers interested in particular proveniences to access the information about a
burial or ceremonial deposit in uncoded form.
This makes the appendix valuable as a quick

340

reference when reading about particular sites

or burials, and as a means of assessing whether
the coding scheme that we developed for
particular variables is appropriate to a specific
study. Second, the provenience sheets contain
a limited amount of information that is not
included among the data base variables or
codes, such as information about atypical
burial characteristics, measurements of certain
artifacts, associations between certain artifacts
and pieces of fabric or other organic materials,
and species names for particular ocean shells.
Third, the provenience sheets are presented as
separate files for each site, making information
for a particular site easily searchable using the
Find function within MS Word. Researchers
wishing to relocate a burial or deposit that
contained a specific artifact or material, such as
the cremation with the large obsidian deposit at
the Hopewell site, or the burials from Hopewell
and Seip that contained copper nostril inserts,
can locate such proveniences in a matter of
seconds. Additionally, researchers interested
in studying specific materials, such as copper,
mica, galena, or pipestone, can search the
provenience sheets for each mention of these
materials to better understand their distribution
within a certain site, across different sites, or
as the medium of particular artifact forms. For
researchers who are interested in searching for
all examples of a given artifact class in all
sites in Ohio and in studying the details of its
various contexts of deposition, all provenience
sheet files can be combined into one long serial
list of proveniences. This global file can then
be efficiently searched for all instances of the
artifact and its contexts.
There are some caveats to consider,
however, when using the provenience sheets.
The provenience sheets were not initially
designed with publication in mind. They were
intended as a tool to assist us in coding of
the data base. Most of the provenience sheets
contain information drawn from the primary
data source, such as a site report or field notes,
for a particular provenience. They may or may
not contain additional information drawn from
field notes, accession records, direct observation of certain artifacts, or publications by

D. TROY CASE AND CHRISTOPHER CARR

other authors who noted errors or inconsistencies in the primary sources. In retrospect,
it would have been ideal to have kept track
of all additions to, and the occasional subtractions from, each provenience sheet beyond its
primary data source, as well as the particular
sources of any new information. However, this
was not done systematically. Sometimes when
updating the data base with information from
the supplemental sources listed above, we added
new information to the provenience sheets with
a note indicating the source, and sometimes we
added the information without a source. For
many sites, when presented with new information from supplemental sources, we simply
bypassed the old provenience sheets and made
additions or changes directly to the data base.
In general, the later a site was coded for
inclusion in the data base, the more likely it
is that the provenience sheets contain exactly
the same information as the data base. Thus,
the greatest discrepancies are most likely to
be found between provenience sheets and the
data base for the sites of Hopewell, Seip, and
Turner. When differences are found between a
provenience sheet and the data base, we place
greatest confidence in the data base. Despite
these departures of the provenience sheets from
the data base, the great bulk of information
in the data base is replicated in the provenience sheets. They remain a very useful tool for
overviewing particular burials and for locating
ones with certain characteristicssomething
that we found repeatedly by direct experience
and that convinced us that they should be
published.

ERROR CHECKS
A number of error checks have been made
on the HOPEBIOARCH data base. Both of
the primary steps in data entrythe transferring
of information from site reports and records
to the text-format provenience sheets, and
the translation of the provenience sheets into
the coded data base of variableshave been
checked.
Three rounds of checking were systematic.
First, coded entries in HOPEBIOARCH for the

DOCUMENTING THE OHIO HOPEWELL MORTUARY RECORD

Seip-Pricer mound, the Ater mound, the Burial

Place in the Great Enclosure of Turner, and
Mound 25 in the Hopewell site were checked
against analogous entries for these sites in data
bases created by Nomi Greber (1976) and
Timothy Lloyd (n.d.). These comparisons span
the kinds, numbers, and materials of grave
goods, as well as bodily variables and tomb
form attributes for individuals. The comparisons are reported in detail in Chapter 14. They
indicate the very good to excellent coverage
of written records and their translation into the
coded HOPEBIOARCH data base, and very
good to excellent inter-observer consistency in
coverage and translation. Thus, the precision
or replicability of the HOPEBIOARCH
data base is known to be high for these
four sites.
In addition, the entirety of the data base
was checked twice for the translation of the textformat provenience sheets into the coded data
base of variables. One check was made when the
data base was nearly finished in 1999, with the
exclusion of a number of sites from primarily
southwestern Ohio. Beau Goldstein compared
each provenience sheet to the data base and
flagged suspected errorsdata believed to be
missing, extra, or mistranslated in the data
base. Goldstein and Case then met to resolve
these discrepancies. Commonly, primary and
secondary sources were revisited in order to
determine whether a change in the data base was
warranted. Occasionally, the discrepancies were
attributable to differences between Goldstein
and Case in how they thought a textual entry in
the provenience sheets should be coded into the
variables of the data base.
Another systematic check of the data base
against the provenience sheets was conducted
in Fall 2006, after all proveniences from all
sites had been coded into the data base. Ashley
Evans, a graduate student in bioarchaeology
at Arizona State University, compared the text
entries in the provenience sheets for the entire
data base against the coded entries in the data
base itself and flagged potential errors. These
potential errors were then evaluated against
primary sources and, as necessary, corrected
by Case.

341

A total, systematic check of the textual

information in the provenience sheets against
the primary sources was not made, other than
indirectly through the comparison of our coded
data base to ones devised by Greber and
Lloyd for select sites (see above). A total
check between original sources and the provenience sheets would have been an impractically
huge jobliterally years of effort. However,
checks on a sample of provenience sheets were
made in the course of checking their consistency with the data base, which sometimes
required going back to primary references, as
described above.
Checks of some provenience sheets against
primary sources were also made in the following
manner while the data base was being built.
Information on the provenience sheets was
first recorded from published site reports. As
field notes, accession records, and field and
artifact photographs were gathered and their
information was added to the provenience
sheets, discrepancies between the published
information as written down on the provenience
sheets and the additional, unpublished sources
were checked by going back to the published
site reports. Sometimes the two sources actually
disagreed, whereas at other times an error
had been made by us in writing the provenience sheets. These comparisons helped to
clarify especially the numbers of artifacts
of particular types present in a burial and
sometimes the forms and types of artifacts,
themselves. We generally found that the error
rate in transferring information from the site
reports to the provenience sheets was gratifyingly low.
Considering all of these several kinds of
error checks, we conclude that the data base
should quite accurately reflect the contents of
the various data sources available to us. The
main source of any errors found in the data
base will probably turn out to have been caused
by occasional misinterpretation of the primary
information sources rather than input errors.
Misinterpretation would be the most likely
source of error because much of the coded information was taken from written descriptions of
artifacts and tombs rather than illustrations or
direct observations of them.

342

CONCLUSION
The HOPEBIOARCH data base brings together
an unprecedented quantity of information on
Hopewell burials from nearly all excavated and
documented mortuary sites in Ohio. Together
with the provenience sheets and maps included
with the book, this data base offers great

D. TROY CASE AND CHRISTOPHER CARR

potential for future research into the lifeways

of Ohio Hopewell people. It is our hope that
easy access to this information will encourage
researchers to delve more deeply into the
Hopewell archaeological record in order to better
understand the varied social and ceremonial ways
and world views of Hopewell peoples across
Ohio.